Spray dust suppression technology from a bibliometric perspective: Research status and trend prediction based on deep learning

Abstract

Spray dust suppression technology plays a critical role in controlling coal mine dust and has attracted growing attention in recent years. However, the diversity of research directions has made it difficult to clearly anticipate future developments in the field. To address this, the present study adopts a bibliometric approach, integrating visualization tools such as VOSviewer, CiteSpace, and Scimago with advanced deep learning models including BERTopic, Holt-Winters, Prophet, and Bi-LSTM. A comprehensive analysis was conducted on relevant publications indexed in the Web of Science Core Collection from 1994 to 2024 to identify research hotspots and forecast future trends. The findings reveal that spray dust suppression research has undergone three distinct phases: initial development, steady growth, and rapid expansion, with a marked increase in research activity after 2017. China, the United States, and Australia are the main contributors, with research concentrated in mining-focused universities and institutes. Keyword co-occurrence networks and BERTopic modeling indicate that current research centers on environmental pollution control, spray fluid dynamics simulation, the application of surfactants and charged mist, spray system optimization, and intelligent dust suppression technologies. By combining burst keyword analysis with multi-model forecasting, the study predicts that future research will emphasize the development of novel eco-friendly materials, multi-technology synergistic enhancements, and the construction of intelligent dust suppression systems. The “bibliometric analysis–topic modeling–trend prediction” methodological framework established in this study provides conceptual support for subsequent research.